Deviating Voice. Representation of Female Characters and Feminist Readings in 1990S Anime 2019

Repositorium für die Medienwissenschaft

Minori Ishida Deviating Voice. Representation of Female Characters and Feminist Readings in 1990s Anime 2019 https://doi.org/10.25969/mediarep/16392

Veröffentlichungsversion / published version Zeitschriftenartikel / journal article

Empfohlene Zitierung / Suggested Citation: Ishida, Minori: Deviating Voice. Representation of Female Characters and Feminist Readings in 1990s Anime. In: IMAGE. Zeitschrift für interdisziplinäre Bildwissenschaft. Special Issue of Issue 29, Jg. 15 (2019), Nr. 1, S. 22– 37. DOI: https://doi.org/10.25969/mediarep/16392.

Erstmalig hier erschienen / Initial publication here: http://www.gib.uni-tuebingen.de/image/ausgaben-3?function=fnArticle&showArticle=517

Nutzungsbedingungen: Terms of use: Dieser Text wird unter einer Deposit-Lizenz (Keine This document is made available under a Deposit License (No Weiterverbreitung - keine Bearbeitung) zur Verfügung gestellt. Redistribution - no modifications). We grant a non-exclusive, Gewährt wird ein nicht exklusives, nicht übertragbares, non-transferable, individual, and limited right for using this persönliches und beschränktes Recht auf Nutzung dieses document. This document is solely intended for your personal, Dokuments. Dieses Dokument ist ausschließlich für non-commercial use. All copies of this documents must retain den persönlichen, nicht-kommerziellen Gebrauch bestimmt. all copyright information and other information regarding legal Auf sämtlichen Kopien dieses Dokuments müssen alle protection. You are not allowed to alter this document in any Urheberrechtshinweise und sonstigen Hinweise auf gesetzlichen way, to copy it for public or commercial purposes, to exhibit the Schutz beibehalten werden. Sie dürfen dieses Dokument document in public, to perform, distribute, or otherwise use the nicht in irgendeiner Weise abändern, noch dürfen Sie document in public. dieses Dokument für öffentliche oder kommerzielle Zwecke By using this particular document, you accept the conditions of vervielfältigen, öffentlich ausstellen, aufführen, vertreiben oder use stated above. anderweitig nutzen. Mit der Verwendung dieses Dokuments erkennen Sie die Nutzungsbedingungen an. Minori Ishida 石田美紀

Deviating Voice. Representation of Female Characters and Feminist Readings in 1990s Anime

Abstract

In the 1990s, Japanese anime sophisticated both their ›visual database‹ and their ›voice database‹ for their character design. These two ›databases‹ usually cooperate in a complementary manner in order to construct characters for an audio-visual medium. In the following article, however, I am going to point out that there are always possibilities of deviation, because, fundamentally, the visual appearance and the voice of the character are created independently. This has, in fact, opened up the possibility to introduce a new style of characters like Haruka Tenou, one of the most popular characters in the Sailor Moon series (1992–1997). According to Azuma Hiroki, moe (affective responses) toward characters had drastically altered the reception of anime in the 1990s, preparing the way for the so-called ›kyara-moe‹. Within otaku (fan) cultures, however, another kind of reception took place, which was inspired by female, queer characters, such as Haruka or her successors. Feminist audiences who experienced moe toward these characters interpreted them enthusiastically: with regard to the gender and the sexuality of the protagonists, they created their own narratives.

Introduction

From birth to death, we are enclosed within our given bodies. Even though we are able to modify our bodies to some extent, we are soon confronted with

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 22 Minori Ishida 石田美紀: Deviating Voice

their limitations. In contrast to us, anime characters are created and modified according to will. Characters are artificial beings. Various opinions have been expressed with regard to their artificiality, which remains the core feature of fictional beings. Most existing studies, including Azuma Hiroki’s theory of ›database consumption‹ (which I will examine later), have focused on the visual aspects of animated characters. However, considering only visual aspects is not enough to understand how characters are constructed and represented in animated works. Since animation must be understood as an audio-visual medium, most of its characters are also given a voice, which plays an important role in animation. If a character’s voice suddenly changes, for instance, we do not think that he or she is the same person we have come to know before. Therefore, an identical voice is usually necessary for the viewers to recognize coherent characters in animation. This article explores the role of voice in animation on the one hand, by analyzing the ways in which it supports character design, as well as the relationship between visual elements and characters’ voices on the other. To this end, I choose—from the vast area of animation—Japanese TV anime series from the 1990s as a site for the following discussion. In the history of anime, the decade of the 1990s is considered to be of paramount importance. In 1963, when televised anime series became popular with the introduction of Tetsu- wan Atomu (Astro Boy), this new genre was perceived strictly as entertainment for children. However, as the anime industry started developing a variety of narratives, the age of the audience expanded. In the 1990s, TV anime series had long reached a young adult audience. In response to the expanded audience, many influential masterpieces targeting young adults were produced, such as Kōkaku Kidōtai (Ghost in the Shell, 1995), Shin Seiki Evangerion (Neon Genesis EVANGELION, 1995–1996), Cowboy Bebop (1998), and many more. By recalling these titles, we can easily recognize that anime was already ›mature‹ as a young-adult genre in the 1990s. Moreover, with regard to audience reception, an outstanding phenom- enon occurred in the 1990s: the young-adult audience became deeply and in- tensely fascinated with anime characters. This attitude revealed a new ten- dency of increased social anime acceptance: the otaku subculture had emerged. So-called ›otaku‹ (devoted fans) began to describe their feelings, which even included sexual desires toward anime characters, with the term ›moe 萌え‹: an intense affective response toward fictional characters. Thus, firmly embedded within various social practices, anime reached its maturity through the public’s fascination with characters. Of course, both with regard to the character and the audience, moe can be evoked regardless of genre. None- theless, the discourse on moe is almost exclusively centered around the reception of female characters by male audiences, as is the case within Azuma’s theory of database consumption (AZUMA 2009). These contexts make us aware of two overlooked issues with regard to female anime characters in the 1990s. One is the contribution of voice to character design. As was already mentioned, anime characters are not only visual

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 23 Minori Ishida 石田美紀: Deviating Voice

entities. Hence, questions with regard to the relationship between their visual elements and their voices arise. In fact, female anime characters exhibit interesting varieties of vocal expression. Another overlooked issue is the reception of female anime characters by female audiences. According to existing studies, otaku have mostly been assumed to be male. However, there is actually a large portion of enthusiastic female fans within anime audiences, which indicates that female fans can also be considered ›otaku‹. The reception of female anime characters by female audiences, and whether this reception is different to male reception, is a research subject that deserves closer attention. To investigate these two issues, I am first going to consider the ›voice database‹ for character designs and re-evaluate the concept of ›database consumption‹ proposed by Azuma. Second, I am going to analyze the female characters from the Sailor Moon series (1992–1997), a representative anime show from the 1990s, by focusing on the relationships between their visual elements and their voices. Third, I will demonstrate that there was a positive reception of female anime characters by their female audiences in the 1990s. I will dis- cuss readings relevant to the artificial and even queer relationships between visual elements and voices in the design of anime characters.

Moe and Database Consumption

In 2001, cultural critic and philosopher Azuma Hiroki published Otaku. Japan’s Database Animals (the English translation referenced in this article was released in 2009). In this book, Azuma analyzed the close relationship between moe and what he coined ›the database‹. According to the author, characters in the 1990s were rapidly rising to prominence within otaku culture; a social ›database‹ of shared assumption and conventions could be identified as a resource for recurring moe-elements, mediated by anime characters. Otaku strongly demand the experience of moe from their reception of fictional characters. Therefore, the narratives to which the characters belong became less and less important. Consequently, only the database is relevant, and otaku- desire seeks out works that function like a database themselves, which allow fans to abstract idealized moe-elements most efficiently. Azuma named this new kind of reception »database consumption« (AZUMA 2009: 47). As a significant example for such ›database consumption‹, Azuma dis- cusses Di Gi Charat, also known as Dejiko. This character was originally (in 1998) created as a mascot for a retail shop chain called Gamers. Gamers spe- cializes in anime, manga, and games. Di Gi Charat, in fact, did not have any narrative background-story at all. Interestingly, however, soon after her introduction, she became very popular within her otaku audience. Subsequently, she also became the heroine of an anime series. Because moe towards characters changes the hierarchy of character traits traditionally governed by narrative, Di Gi Charat rose to popularity without or before being part of any story: she therefore existed before any narrative on which her ›existence‹ could be

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 24 Minori Ishida 石田美紀: Deviating Voice

based. In the 1990s, characters could thus become independent from stories, and their abilities to ›fascinate‹ the audience became a powerful hub for the otaku-imagination. Azuma moreover points out that every design feature and character trait of Di Gi Charat was based on popular moe-elements from the otaku database, which explains why she became so popular in a short span of time. He notes: However, one cannot quite say that the design was particularly original or attractive. In fact, the design of Digiko [sic] is a result of sampling and combining popular elements from recent otaku culture, as if to downplay the authorship of the designer (AZUMA 2009: 42).

In Azuma’s discussion, his notion of ›database‹ indicates that the Japanese society is increasingly governed by postmodern conditions. The acceleration of information technology had finally brought about the ›death of the author‹ in Japan, as well as the decline of all ›grand narratives‹. In their wake, otaku and their mode of database consumption have emerged. However, if we change our point of view slightly, we can recognize many positive aspects of database consumption. The database then appears as a mere foundation, supporting both the creators and the audience. More importantly, database consumption allows for a free interpretation of characters by the audience. This freedom will be discussed in more detail later.

The ›Voice Database‹ of Anime Character

One aspect is of particular importance in Azuma’s lucid conception of the relationship between moe and character design. Although he does point out briefly that some instances of Di Gi Charat are represented to talk with the particle »~nyo« (a cat-like sound), which is considered a moe-element (cf. AZUMA 2009: 47), the rest of the moe-elements are purely visual ones. In fact, Azuma’s database is essentially a visual database. Therefore, his explanation cannot ade- quately describe characters developed within audio-visual media, because the anime-database should also include voices. Historically, anime has depended extensively on characters’ voices. I will briefly demonstrate the functions of anime voices by referring to the first TV-anime series: Astro Boy (1963). Right in the first episode, the inventor Doc- tor Tenma holds his dead son in his arms—which he should go on to recon- struct as the robotic hero of the series—and he screams in a heartfelt way. At that moment, nothing on the screen is moving or animated. His overall expression is completely different from the original manga version, however, because we come to know the character’s deep suffering through his voice. It is this very cry that establishes the whole tragedy between father and son. Moreover, this scene demonstrates unambiguously that voice can give internal depth to the characters at the same time. Furthermore, anime voices can also drive the narrative as a whole, particularly when movement is limited (cf. GAN 2009: 302).

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 25 Minori Ishida 石田美紀: Deviating Voice

In Japan, animation has traditionally emphasized movement, as it was established in a quite different style than Disney’s animation in the United States. We must nevertheless remember that the maturation of the anime system was inevitably linked to the maturation of voice acting as one of its most defining characteristics. Nowadays, more than 60 years after the TV-broadcast of Astro Boy, the anime audience (including the otaku) can imagine a character’s appearance and design just by listening to his or her voice. The opposite can also be true: the audience can imagine a certain type of voice just by looking at the character’s visual design. Tone, pitch, and voice articulation, in association with the visual elements, allow us to conceive a character’s personality as, for instance, cool, clumsy, or tsundere (girls that appear cold or even hostile in public, but are revealed to be lovely and caring in private), and so on. Accordingly, the voices of characters could also be classified into patterns of moe-elements, similar to Azuma’s ›cat ears‹ or ›maid costumes‹ as visual traits. Moe-elements of character voices have thus entered the database of fans, where they are available as another kind of resource for the audience’s desire.

The Voice of Di Gi Charat

To illustrate this more concretely, I will return to Di Gi Charat whose anime series was released in 1999. Di Gi Charat is in fact an interesting example for a discussion of character voices. The audition to cast the voice actress for Di Gi Charat was held publicly, and six finalists were selected. Watching the footage of this audition, it is impossible to ignore that all the finalists’ voices were strik- ingly similar.1 Of course, each finalist performed Di Gi Charat in their own way. However, the tone and pitch had many common characteristics. More importantly, the style of the voice acting was not at all strange or unfamiliar to an audience of anime fans. Listening to similar voices, they could easily imagine a female character of specific traits (that must be ›pretty‹, for instance). What is more, the anime audience—including the finalists and judges—could be said to expect a certain tone, pitch, and voice articulation in accordance with Di Gi Charat’s visual elements (props such as ring, or a specific dress). Through these interconnections of visual appearances and voices, we can deduct that a voice database must exist for characters and that it is shared by both the anime

1 The footage of the audition was distributed as a special treat for the customers of Gamers, cf. Karinto: Neko no me girarin. Diary, May 17, 2006, http://d.hatena.ne.jp/karinto/20060517/p4 [accessed August 25, 2018]. It can be found on several video-sharing platforms on the internet nowadays, cf. Di Gi Charat seiyū ōdishon 1/2. Nico Nico, October 20, 2008, http://www.nicovideo.jp/watch/sm4989445 [accessed August 26, 2018]; Di Gi Charat seiyū ōdishon 2/2. Nico Nico, October 20, 2008, http://www.nicovideo.jp/watch/sm4989744 [accessed August 25, 2018]. The footage is known for rare recordings of the young Sawashiro Miyuki who became a popular voice actress later. Sawashiro, who was just a junior high school student at the time, participated in the audition and received a special award. Based the popularity of this footage and the number of comments it received, we can conclude that the audience is highly interested in voice actors.

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 26 Minori Ishida 石田美紀: Deviating Voice

audience and the voice actors. This database of vocal features allows them to imagine that Di Gi Charat has a ›pretty‹ appearance, for instance. Characters’ voices, however, are not integrated into the visual database, because their voices are given to them by actors and actresses, whereas the visual elements are drawn by designers and animators. These separate production contexts indicate that anime characters have at least two different com- ponents: visual appearances and voices. In this regard, their bodies are onto- logically different from ours. Therefore, even though a voice can possess moe- elements, its database must exist independently from the visual database. Hence, I will call it the ›voice database‹. Anime characters are constructed from both a visual and a vocal database.

Coherence Between the Visual and the Vocal Database

How does coherence between the visual database and the voice database oc- cur? Typically, these two databases are in close cooperation to produce characters with a high level of consistency. As the Di Gi Charat audition reveals, their tight interconnection can provide easy access to moe-affects. To activate the voice database effectively, voice actors specializing in anime speech patterns are required. Kobayashi Shō, who closely investigated the history of anime voice acting (2015), considers the 1990s as an important era because, during that time, a fandom around anime voice actors (lending their voices exclusively to anime) was established. Before then, many voice actors first made a debut as stage actors, performed on theater and radio plays, or dubbed foreign movies and TV dramas. For such actors, anime was only one part of their jobs (cf. KOBAYASHI 2015: 12). Yokomori Hisashi, for example, the voice actor of Doctor Tenma in Astro Boy, mainly appeared in live action films. His acting career was thus not only supported by anime. Specialized anime voice actors internalize a set of anime production conventions right from their debut on. Amongst these conventions, the most important one is that, in anime, voice recording is conducted at the latest possible stage of the production. Anime voices are typically added to more or less finalized character animations. In Disney animation, in contrast, voices are mostly recorded first. The moving images are then produced to match the existing tracks. For this reason, in anime, the visual appearance of the characters provides a stronger anchoring for the voice acting in comparison to Disney animations. It is therefore not at all peculiar that the visual and the vocal databases cooperate seamlessly in order to create a conventional consistency for anime characters. Since the 1990s, when anime voice actors were at their hey- day, the two databases were consequently employed in cooperation to each other to contribute to the production of moe.

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 27 Minori Ishida 石田美紀: Deviating Voice

Voice Deviation from the Visual Database in the 1990s: Sailor Uranus

As was shown by the audition of the Di Gi Charat anime series, the voice database and the visual database cooperate seamlessly in most cases. I would like to emphasize, however, that there are always possibilities of deviation, since the visual appearances and the voices of characters are created independently from each other. Occasionally, the voice can deviate from the visual elements, in what must be an unconventional combination of the two databases. To illustrate this point, I will take a closer look at an anime from the 1990s, namely the Sailor Moon series mentioned before. The story revolves around themes of love and friendship among girl-heroines: schoolgirls who acquire the special ability to transform into sailor-warriors, fight against evil, and form great friendships. Although this series was originally aimed at young female recipi- ents, it became highly popular within a wider audience, including otaku, and had a considerable effect on the Japanese society at that time. Moreover, Sailor Moon is particularly suggestive and important for discussing the relationship between the visual and the vocal database. Hence, I will analyze Haruka Tenou, also known as Sailor Uranus, who appears in the third season of Sailor Moon S (1994–1995), since this character exhibits a radical tension between the two databases.

Fig.1: Haruka’s first appearance in Sailor Moon S (episode 92, 1994)

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 28 Minori Ishida 石田美紀: Deviating Voice

The protagonist Usagi, also known as Sailor Moon, and her friend Minako, also known as Sailor Venus, are playing a racing game at a game center in the series’ 92nd episode. When they start bickering, a person unknown to them (Ha- ruka) approaches them, saying: »Hi girls, may I drive with you?« (translation M.I.). Her voice is lower and huskier than Usagi’s and Minako’s, her hair is short, and she is dressed in the male uniform of a Japanese high school student: she is wearing trousers and a tie. Therefore, at her first appearance, Ha- ruka looks decidedly like a high school boy. Moreover, Haruka refers to herself as »boku«. In Japanese, ›boku‹ is a first-person pronoun for male speakers. Taken together, the vocal database reinforces the visual database to present Haruka as a young boy. Not surprisingly, Usagi and Minako fall in love at first sight with Haruka (cf. fig. 1).

Fig. 2: Sailor Uranus in a battle scene of Sailor Moon S (episode 92, 1994)

However, in a subsequent battle scene (a recurring feature of the Sailor Moon series), the coherence between the two databases is radically altered. When Haruka transforms to Sailor Uranus, she wears a mini skirt, as well as a tiara on her forehead. Her costume is as feminine as the dresses of the other Sailor warriors (cf. fig. 2). At the same time, however, Haruka often shouts with a voice that is lower-pitched and more powerful than in the daily scenes of her usual self. At moments such as these, the visual elements (uniform-like dresses with mini skirts) and the vocal elements (a powerful roaring)—signifying femininity and masculinity, respectively—coexist within the same character. Be- cause the voice deviates from the visual elements, the discrepancy between the two databases becomes evident. Consequently, Haruka’s gender remains

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 29 Minori Ishida 石田美紀: Deviating Voice

undefined and ambiguous. After the battle, a more intricate and interesting sit- uation occurs with regard to Haruka’s sexuality and gender. »I never said that I am male«, Haruka admits to Usagi and her friends. At this moment, Haruka’s sex is clearly identified as female. Her gender, however, remains a much more complicated issue. In comparison to the voice she employed during battle, she is speaking in a slightly higher pitch later, while at the same time continuing to call herself »boku« (the first-person pronoun for male speakers). Additionally, if we pay attention to the visual elements, we can also observe minor changes in her body: she is not wearing the tie, her chest is drawn somewhat roundly. Compared to the battle scene, her eyes are also depicted in larger sizes; in other words, her femininity is in fact emphasized. Thus, the discrepancies and ten- sions between the visual and the vocal database is uphold without being con- solidated. Consequently, Haruka continues to cross the gender boundary; this is only possible by the artificiality of characters that are always subject to a double-system of visual appearances and voices.

Fig. 3: Haruka refers to herself by the pronoun »ore« (TAKEUCHI 1994/1: 171)

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 30 Minori Ishida 石田美紀: Deviating Voice

Fig. 4: Sailor Uranus calls herself »atashi« (TAKEUCHI 1994/2: 61)

Before discussing more closely what this playful and provocative deviation in- itiated, I want to stress that this depiction of Haruka’s complex gender is specific to the anime version of Sailor Moon, even in striking contrast to the Haruka of the manga version. In the Sailor Moon-manga (created by Takeuchi Naoko, 1991–1997), Haruka’s appearance is almost the same as in the anime. However, there are important differences between the two versions. In the manga, Ha- ruka chooses ›ore‹ as her first-person pronoun in daily life (cf. TAKEUCHI 1994/1: 171, fig. 3). In Japanese, ›ore‹ is an even much more masculine pronoun than ›boku‹. Here, the verbal expression closely corresponds to the male attire. Ha- ruka appears as male in daily life, but she wears a female costume during battle—just as in the anime version. Interestingly, however, Haruka changes her first-person pronoun again to »atashi« during fight scenes (cf. TAKEUCHI 1994/2: 61, fig. 4), which is a first person pronoun for female speakers in Japanese. In the manga, the correspondence between the verbal expression and the costume is thus quite static and even rigid. The masculine verbal expression is adopted with the male costume, while the feminine verbal expression is adopted with the female costume. Therefore, in the manga, Haruka never actually disturbs the gender boundary. Moreover, the manga-Haruka transforms to Sailor Uranus and speaks to Usagi about her gender during one specific battle. There, it is explained that Sailor Uranus is in fact male and female, a warrior with both powers (cf. TAKEUCHI 1994/3: 20, fig. 5). The Manga version concludes by informing us that Sailor Uranus is, in fact, androgynous.

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 31 Minori Ishida 石田美紀: Deviating Voice

Fig. 5: Sailor Uranus reveals his/her sexual identity to be androgynous (TAKEUCHI 1994/3: 20)

Manga and anime are different media. In manga, there is no possibility for a direct vocal expression, so the presentation of Haruka is inevitably different to her anime counterpart. However, if the anime version had been more loyal to the original—which defines its gender by corresponding first-person pronouns and costumes—the anime-Haruka would have spoken differently: when Sailor Uranus said »atashi«, the voice would have betrayed a higher pitch, it could have sounded much more feminine. In comparison with the androgynous manga-Haruka, her anime version crosses the gender boundary and stands out significantly, since the identification of Haruka’s gender is not an important issue in the anime version. While Haruka’s gender identity is the subject of six episodes in the manga, the same issue is resolved within a single anime-episode. But while Haruka’s gender is only a minor explicit topic within the anime narrative, these questions are presented as much more complicated in formal terms. In the episodes that follow, she stays in male attire and speaks in a much lower-pitched voice (even in comparison to the only male protagonist, Ma- moru), although she is definitively presented as female. For this reason, we can say that she deviates from stereotypical female characters to a large degree. Switching our attention towards Haruka’s sexuality, we also need to take Michiru into account, also known as Sailor Neptune. The two of them have a provocative relationship. Because the narrative informs us that both of them are female, they are understood to have a lesbian relationship. More importantly, however, Usagi and the other characters respect them just as they are. Thus, there is no denial of sexual diversity in Sailor Moon (cf. episode 95), which is one of the most positive and important achievements of the anime series. The playful discrepancy of the vocal and the visual database highlights the artificiality of anime characters. As has been shown with respect to Haruka,

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 32 Minori Ishida 石田美紀: Deviating Voice

such discrepancies between visual appearance and voice can expand the repertory of female characters. This opens up the possibility of introducing a new type of characters.

Deviance as an Indicator of Maturity in the Character Creation of the 1990s

How can we evaluate the discrepancies between the visual and the vocal database behind Haruka’s construction? We could think of it as a kind of Brechtian ›alienation‹-device, criticizing conventions. However, I do not think that this ap- plies to Haruka for two reasons. The first reason lies in the context of Haruka’s creation. She appeared two years after the first season had started. In fact, she is the eighth warrior within the show. Hence, it was necessary for the creators to somehow evade the preceding seven Sailor warriors’ characteristics. At the time, ›adult‹ and ›masculine‹, themes had not yet been used. The discrepancy under discussion was thus probably chosen as nothing more than a defining characteristic of a new character. In fact, Haruka is older than Usagi and her friends and she is also presented as a genius driver. In the Japan of the 1990s, owning a driver license was considered a symbol of the matured male. Thus, the deviation from stereotypical female characteristics occurred mainly for the purpose of immersing the audience into the story, by using a fresh and attractive character. The second reason why I am hesitant to consider Haruka as a critical intervention relates to the vocal database itself. The role of Haruka was performed by voice actress Ogata Megumi, who gave Haruka a persuasive and attractive masculine voice. At first glance, the fact that the masculine voice was performed by a voice actress could be seen as yet another challenge for norms of gender and sexuality. It was not an actual deviation, however, because it had already been an established convention within the voice database itself. Let us return to Astro Boy in 1963. Atomu, the eponymous boy protagonist, was voice-acted by an actress (Shimizu Mari) even then. Since this pioneering per- formance, many boys’ roles have been voiced by actresses, regardless of genre. This type of casting thus quickly established a firm convention. Before Haruka, Ogata herself had performed as the high school boy Kurama in the anime series Yū Yū Hakusho (1992–1995). After Haruka, she continued her work as the junior high school boy Shinji Ikari in EVANGELION. Thus, Haruka’s casting followed an established convention after all. At the same time, I want to point out that Haruka’s voice reopened and subverted that convention from inside the vocal database. In her case, a boy’s voice—performed by a voice actress—is given back to a female character. The discrepancy in the casting thus occurred from the inside of the vocal database: it was ›corrected‹ during the audience’s first encounter with Haruka, right after she appeared to be a boy. In the following battle sequence, however, the vocal database conflicts drastically with the visual one, because Haruka’s feminine

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 33 Minori Ishida 石田美紀: Deviating Voice

costume amplifies and emphasizes the deviation between the two. Within this interplay of discrepancies and deviations, Haruka was born. Her introduction renewed the visual and the vocal databases for anime characters. Continuing to disturb gender and sexual norms, Haruka converted the artificiality of anime characters into the embodiment of gender and sexual diversity. As an achievement of 1990s anime, the queerness embodied in Haruka reminds us of Judith Butler. In criticizing the repressiveness of gender, sex types, and immobilization, Butler asked the following question: »What constitutes the possibility of effective inversion, subversion, or displacement within the terms of a constructed identity? What possibilities exist by virtue of constructed character of sex and gender?« (BUTLER 1990: 32, original emphasis). Haruka, an artificially constructed character, responds to this question unexpectedly, but fully. However, I am going to emphasize that such possibilities of effective inversion must be searched outside of Butler’s discourse, with respect to the process of anime character creation and development. Creating a new and attractive character can be seen as a series of trial-and-errors, consisting of interrelations between vocal and visual databases.

New Tendencies after Haruka

Does Haruka stand alone in this history of anime? In the remains of this article, I will describe what happened after she was introduced to the genre. Connected to this provocative character is the anime director Ikuhara Kunihiko, the chief director of Sailor Moon S. His interests often revolve around questions of gender and sexuality. But, even after he left the Toei-animation studio (responsible for the Sailor Moon anime), discrepancies between the visual and the vocal database could be found within the series. In the last season (Sailor Moon Sailor Stars, 1996–1997), three new male protagonists, members of the boy band »Three Lights«, appeared. Their voice was performed by voice actresses—again, according to the prevailing casting convention. In scenes de- picting their daily life, these characters were evidently shown as boys, both in terms of visual appearance and voice. During battle scenes, however, they transformed to female warriors wearing sexy and feminine clothes. Because they also relied on discrepancies between the visual and the vocal database, they can be considered Haruka’s legitimate successors. In these ways, the Sailor Moon series introduced new types of characters to anime by using both databases in innovative ways. After the series ended, playful discrepancies between visual appearances and voices could be found in many other anime. Let us briefly consider another example from Ikuhara’s works. Utena, a heroine of Shōjo Kakumei Utena (Revolutionary Girl Utena, 1997) calls himself »boku« (similar to Haruka) and wants to be a prince. There is a long tradition for this in manga and anime as well. The visual appearance of Oscar in Berusaiyu no Bara (The Rose of Versailles, the manga was serialized in 1972–1973, the anime aired in 1979–1980) can be described as

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 34 Minori Ishida 石田美紀: Deviating Voice

that of a beautiful lady in male attire. However, Utena’s combination of feminine, pink, and long hair with remodeled student clothes deviates from Oscar’s typical male attire. More interestingly, she does not have a low-pitched voice, but rather a high-pitched and cheerful one. Ikuhara created a new type of character by slightly modifying the regular interplay of the vocal and the visual database. Departing from previous deviations (Haruka or Oscar) to a new one, Utena embodies the personality of a girl who searches for her own way to express her gender and sexuality.

From Kyara-Moe to Feminist Readings in the 1990s

So far, I have described playful deviations between the vocal and the visual database in 1990s anime. Coming to a close, I am going to consider some aspects of audience responses to these deviations. Again, we can draw insights from Azuma’s ›database consumption‹. Azuma states that EVANGELION functions essentially not as a narrative, but rather as an assemblage of information that drives the otaku desire for moe, which the otaku call ›kyara-moe‹ (or, in the original translation, »chara-moe«, AZUMA 2009: 36). Kyara-moe refers to a reception of anime focusing exclusively on characters, which is an effect of database consumption. Although Azuma does not evaluate database consumption in great detail, he describes the term ›database‹ as follows: Therefore, as these databases display various expressions depending on differing modes of interpretation by users and consumers, once consumers are able to possess the settings, they can produce any number of derivative works that differ from the originals (AZUMA 2009: 32).

Interestingly, the database can eventually become an apparatus to provide the audience with freedom of interpretation. It is thus necessary to re-examine the reception of the Sailor Moon series in the 1990s. It is famous, if not notorious for its enthusiastic otaku-fans, mainly because their fan-creations manifest moe for the characters. The Sailor Moon reception can be seen as typical for kyara-moe, resembling the reception of EVANGELION and other series. In fact, each sailor warrior’s characterization corresponds exactly to affective moe-elements in Azuma’s visual database. For example, Usagi is a typical ›clumsy but adorable girl‹; Ami is a typical ›serious and honorable student‹, Rei is a typical ›shrine maiden‹. However, the Sailor Moon series also created a new type, Ha- ruka, which exploits discrepancies between the visual and the vocal database. Moreover, other characters such as the »Three Lights« and Utena followed in her wake. All of these characters are thus deviations from the conventional coherence between the two databases. Interestingly enough, many reviews and essays by feminist writers have been published in the 1990s which focus exactly on these kinds of characters. In 1997, for instance, Yamaguchi Kayoko described Haruka’s unique- ness, comparing her to Oscar from The Rose of Versailles, as follows:

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 35 Minori Ishida 石田美紀: Deviating Voice

Even the characters who seem to be male need a female body in order to participate in the public sphere within the narrative. This means that this story can realize a new world, where the gender bias will be reduced as much as possible, and new relationships will emerge between men and women that have never existed before (YAMAGUCHI 1997: 75, translation M.I.).2

Yamaguchi does not praise everything about the Sailor Moon series. Never- theless, she interprets some of its characters—who continue to cross the gender boundary—to provide resources for relationships that might become free from the bias of gender. In the following year of 1998, the anthology Shōjo- tachi no senreki. »Ribon no Kishi« kara »Shojokakumei Utena« made (The Girl’s War History. From »Princess Knight« to »Revolutionary Girl Utena«) was published.3 Most of its contributors enthusiastically discussed Haruka and her suc- cessor, Utena, with regard to themes of gender and sexuality. According to Azuma, the affective potential of moe had drastically changed the reception of anime and its characters in the 1990s, leading to the predominance of kyara- moe. Alongside the otaku culture, however, another kind of reception—inspired by female characters such as Haruka and her successors—took place. The feminist authors in the collection above, who might have experienced moe towards these characters as well, interpreted them enthusiastically and created their own narratives with regard to their gender and sexuality. Each essay or review consequently examined gender and sexual diversity outside of regular norms and conventions. The blossoming of the feminist readings in the 1990s proves that characters, constructed artificially against the backdrop of vocal and visual databases, can symbolize possibilities for freedom and diversity for human beings. After all, we are not only confined by our material bodies, but also by social norms and conventions.

References

AZUMA, HIROKI 東浩紀: Dōbutsuka suru posutomodan. Otaku kara mita nihon shakai. 動物化するポストモダン―オタクから見た日本社会. Tokyo [Kōdan-sha] 2001 AZUMA, HIROKI: Otaku. Japan’s Database Animals. Translated by Jonathan E. Abel and Shion Kono. Minneapolis [Minnesota UP] 2009 BUTLER, JUDITH: Gender Trouble. Feminism and the Subversion of Identity. New York [Routledge] 1990

2 Original: »一見男にしか見えないキャラクターも「公」領域ではすべて女の肉体を持つ世界―それはすなわち、生まれ持った性差というものが極度に矮小化された世界である。そこには、現実社会に存在するどのような男女関係も重ね合わせることが出来ない画期的なドラマを描きだせる可能性が存在していた«. 3 Saitō Tamaki’s essay »Genealogy of Beautiful Fighting Girls«, first published within the same anthology and later (2000) released as Sentō bishōjo no seishin bunseki (English: Beautiful Fighting Girl, 2011), is praised as a classic work with regard to otaku and their sexuality. Considering the amount of critical attention the Sailor Moon series had received, it can be said that its impact on the Japanese society was indeed profound.

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 36 Minori Ishida 石田美紀: Deviating Voice

GAN, SHEUO HUI: Selekutibu animēshon toiu gainen gihō. ›Limiteddo animēshon‹ no genkai wo koete. セレクティブ・アニメーションという概念技法―「リミテッド・アニメーション」の限界を超えて. In: KATŌ, MIKIRO 加藤幹郎 (ed.): Animēshon no eigagaku. アニメーションの映画学. Kyoto [Rinsen Shoten] 2009, pp. 245–308 KOBAYASHI, SHŌ 小林翔: Seiyū shiron. ›Anime būmu‹ ni miru shokugyō seiyū no tenkanten. 声優試論―「アニメブーム」に見る職業声優の転換点. In: Animēshon kenkyū. アニメーション研究, 16(2), 2015, pp. 3–14 SAITŌ, TAMAKI 斎藤環 (ed.): Shōjo-tachi no senreki »Ribon no kishi« kara »Shōjo kakumei Utena« made. 少女たちの戦歴―『リボンの騎士』から『少女革命ウテナ』まで. Tokyo [Seikyu-sha] 1998 SAITŌ, TAMAKI 斎藤環: Sentō bishōjo no seishin bunseki. 戦闘美少女の精神分析. Tokyo [Ota Shuppan] 2000 SAITŌ, TAMAKI: Beautiful Fighting Girl. Translated by Keith Vincent and Dawn Lawson. Minneapolis [U of Minnesota P] 2011 TAKEUCHI, NAOKO 武内直子: Bishojo senshi Sailor Moon. 美少女戦士セーラームーン, 7, Tokyo [Kōdan-sha] 1994 TAKEUCHI, NAOKO 武内直子: Bishojo senshi Sailor Moon. 美少女戦士セーラームーン, 8, Tokyo [Kōdan-sha] 1994 TAKEUCHI, NAOKO 武内直子: Bishojo senshi Sailor Moon. 美少女戦士セーラームーン, 9, Tokyo [Kōdan-sha] 1994 YAMAGUCHI, KAYOKO 山口佳代子: Danso suru ›Bishōjo senshi‹. Iseisō no kyarakutā kara miru anime »Sailor Moon«. 男装する「美少女戦士」―異性装のキャラクターから見るアニメ『セ-ラ-ム-ン』. In: Josei gaku nenpo. 女性学年報, 18, 1997, pp. 71–77

IMAGE | Issue 29 | Special Issue Recontextualizing Characters | 01/2019 37